1 |
In a previous thread, I learned about keeping world simple with |
2 |
--oneshot. I realized how mine had gotten so bloated -- when I |
3 |
update, I edit the --pretend output and feed that directly into |
4 |
emerge without the benefit of --oneshot. |
5 |
|
6 |
So today I started a cleanup project. I began by moving world to |
7 |
world-bloated and running emerge --depclean -p just to see what would |
8 |
happen. The answer is ... a loop! |
9 |
|
10 |
There were a couple of missing or out of date packages and I emerged |
11 |
them. But libusb has to be 10.6 to make some packages happy and 10.7 |
12 |
to satisfy others. |
13 |
|
14 |
I have been down this route before. I don't feel like unmerging |
15 |
either side of the mess, and even if I didn't want the packages, it is |
16 |
way too much hassle to unmerge them one by one as the list of unhappy |
17 |
packages grows. |
18 |
|
19 |
So, what is the proper way to recreate a proper world file? If |
20 |
depclean can finally run one of these days when gentoo gets back in |
21 |
sync, is staring with an empty world file as good as anything else? |
22 |
The idea of trying to make intelligent guesses about which packages |
23 |
are truly top level, out of 3000+ packages, is not enticing. |
24 |
|
25 |
-- |
26 |
... _._. ._ ._. . _._. ._. ___ .__ ._. . .__. ._ .. ._. |
27 |
Felix Finch: scarecrow repairman & rocket surgeon / felix@×××××××.com |
28 |
GPG = E987 4493 C860 246C 3B1E 6477 7838 76E9 182E 8151 ITAR license #4933 |
29 |
I've found a solution to Fermat's Last Theorem but I see I've run out of room o |